复杂背景下红外静脉图像的分割与增强

Segmentation and Enhancement of the Infrared Vein Image in Complex Background

针对医护人员在对患者进行静脉穿刺过程中频繁发生静脉误识别问题,提出了一种红外手掌静脉识别系统。该系统根据近红外光的特性搭建图像采集平台获取原始图像,利用二维最大熵分割法结合区域标记法得到复杂背景中的目标区域,采用反锐化掩模法对目标区域进行增强处理。试验结果表明,目标区域可以从复杂背景中较好地分割出来,提高了图像的对比度,可达到实现红外图像分割和增强的目的。

A new approach is proposed that solve vein erroneous identification problem when doctors puncture patients＇ vein. First, image acquisition platform is built according to the characteristics of infrared light. Then, target area is acquired by algorithm of two - dimensional maximum entropy combined with region labeling algorithm is used to finish image enhancement. Experimental results demonstrate the target area from the complex background and contrast is enhanced. The purpose of infrared image segmentation can be achieved. and unsharp mask can be segmented and enhancement